// 添加请求到队列
func (self *scheduler) Push(req *context.Request) {
// 初始化该蜘蛛的队列
spiderId, ok := req.GetSpiderId()
if !ok {
return
}
self.RLock()
defer self.RUnlock()
if self.status == status.STOP {
return
}
// 当req不可重复时，有重复则返回
if !req.GetDuplicatable() && self.Deduplicate(req.GetUrl()+req.GetMethod()) {
return
}
// 初始化该蜘蛛下该优先级队列
priority := req.GetPriority()
if !self.foundPriority(spiderId, priority) {
self.addPriority(spiderId, priority)
}
defer func() {
recover()
}()
// 添加请求到队列
self.queue[spiderId][priority] = append(self.queue[spiderId][priority], req)
}
func (self *Surfer) Download(cReq *context.Request) *context.Response {
cResp := context.NewResponse(nil)
var resp *http.Response
var err error
switch cReq.GetDownloaderID() {
case SURF_ID:
resp, err = self.surf.Download(cReq)
case PHANTOM_ID:
resp, err = self.phantom.Download(cReq)
}
if resp != nil {
// 确保Response与Request中的Url字符串相等
resp.Request.URL, _ = url.Parse(cReq.GetUrl())
}
cResp.SetRequest(cReq)
cResp.SetResponse(resp)
cResp.SetError(err)
return cResp
}
func (self *Surfer) Download(cReq *context.Request) *context.Response {
cResp := context.NewResponse(nil)
resp, err := self.download.Download(cReq.GetMethod(), cReq.GetUrl(), cReq.GetReferer(), cReq.GetPostData(), cReq.GetHeader(), cReq.GetCookies())
cResp.SetRequest(cReq)
cResp.SetResponse(resp)
cResp.SetError(err)
return cResp
}
func (self *Matrix) SetFailure(req *context.Request) bool {
self.Lock()
defer self.Unlock()
unique := makeUnique(req)
if _, ok := self.failures[unique]; !ok {
// 首次失败时，在任务队列末尾重新执行一次
self.failures[unique] = req
logs.Log.Informational(" * + 失败请求: [%v]\n", req.GetUrl())
return true
}
// 失败两次后，加入历史失败记录
UpsertFailure(req)
return false
}
// 添加请求到队列
func (self *scheduler) Push(req *context.Request) {
self.pushMutex.Lock()
defer self.pushMutex.Unlock()
if self.status == status.STOP {
return
}
// 当req不可重复时，有重复则返回
if !req.GetDuplicatable() && self.Deduplicate(req.GetUrl()+req.GetMethod()) {
return
}
self.SrcManage.Push(req)
}
// core processer
func (self *crawler) Process(req *context.Request) {
defer func() {
if err := recover(); err != nil { // do not affect other
logs.Log.Error(" * Process panic: %v", err)
}
}()
// logs.Log.Debug("**************断点 1 ***********")
// download page
resp := self.Downloader.Download(req)
// logs.Log.Debug("**************断点 2 ***********")
// if fail do not need process
if resp.GetError() != nil {
// 取消该请求的去重样本
scheduler.Sdl.DelDeduplication(req.GetUrl() + req.GetMethod())
logs.Log.Error(" * %v", resp.GetError())
// 统计下载失败的页数
cache.PageFailCount()
return
}
// logs.Log.Debug("**************断点 3 ***********")
// 过程处理，提炼数据
self.Spider.ExecParse(resp)
// logs.Log.Debug("**************断点 5 ***********")
// 该条请求文本结果存入pipeline
for _, data := range resp.GetItems() {
self.Pipeline.CollectData(
resp.GetRuleName(), //DataCell.RuleName
data, //DataCell.Data
resp.GetUrl(), //DataCell.Url
resp.GetReferer(), //DataCell.ParentUrl
time.Now().Format("2006-01-02 15:04:05"),
)
}
// 该条请求文件结果存入pipeline
for _, img := range resp.GetFiles() {
self.Pipeline.CollectFile(
resp.GetRuleName(),
img["Name"].(string),
img["Body"].(io.ReadCloser),
)
}
// logs.Log.Debug("**************断点 end ***********")
}
func (self *Surfer) Download(cReq *context.Request) *context.Response {
cResp := context.NewResponse(nil)
resp, err := self.download.Download(cReq.GetMethod(), cReq.GetUrl(), cReq.GetReferer(), cReq.GetPostData(), cReq.GetHeader(), cReq.GetCookies())
cResp.SetRequest(cReq)
cResp.SetResponse(resp)
if err != nil {
logs.Log.Error(" * %v", err)
// cResp.SetStatus(false, err.Error())
// return cResp
}
cResp.SetStatus(true, "")
return cResp
}
// 添加请求到队列
func (self *scheduler) Push(req *context.Request) {
pushMutex.Lock()
defer func() {
pushMutex.Unlock()
}()
if self.status == status.STOP {
return
}
// 有重复则返回
if self.Compare(req.GetUrl() + req.GetMethod()) {
return
}
// 留作未来分发请求用
// if pholcus.Self.GetRunMode() == config.SERVER || req.CanOutsource() {
// return
// }
self.SrcManage.Push(req)
}
// core processer
func (self *crawler) Process(req *context.Request) {
defer func() {
if err := recover(); err != nil {
// do not affect other
scheduler.Sdl.DelDeduplication(req.GetUrl() + req.GetMethod())
// 统计失败数
cache.PageFailCount()
// 提示错误
logs.Log.Error(" * Fail [process panic]: %v", err)
}
}()
// download page
resp := self.Downloader.Download(req)
// if fail do not need process
if resp.GetError() != nil {
// 删除该请求的去重样本
scheduler.Sdl.DelDeduplication(req.GetUrl() + req.GetMethod())
// 统计失败数
cache.PageFailCount()
// 提示错误
logs.Log.Error(" * Fail [download]: %v", resp.GetError())
return
}
// 过程处理，提炼数据
spider.NewContext(self.Spider, resp).Parse(resp.GetRuleName())
// 统计成功页数
cache.PageSuccCount()
// 提示抓取成功
logs.Log.Informational(" * Success: %v", req.GetUrl())
// 该条请求文本结果存入pipeline
for _, data := range resp.GetItems() {
self.Pipeline.CollectData(
resp.GetRuleName(), //DataCell.RuleName
data, //DataCell.Data
resp.GetUrl(), //DataCell.Url
resp.GetReferer(), //DataCell.ParentUrl
time.Now().Format("2006-01-02 15:04:05"),
)
}
// 该条请求文件结果存入pipeline
for _, img := range resp.GetFiles() {
self.Pipeline.CollectFile(
resp.GetRuleName(),
img["Name"].(string),
img["Body"].(io.ReadCloser),
)
}
}
func makeUnique(req *context.Request) string {
return util.MakeUnique(req.GetUrl() + req.GetMethod())
}